JPH08233668A - Optical fiber type distributed temperature sensor - Google Patents

Abstract

PURPOSE: To provide an inexpensive optical fiber distributed temperature sensor which has high temperature accuracy and distance resolution. CONSTITUTION: A leveling processing device of an optical fiber distributed temperature sensor is composed of an A/D converter 61 and plural sets of adding circuits 600 to process its outputs in parallel to each other, and processing time of the respective sets of adding circuits 600 is set in time of set number times a sampling time interval Ts, and the A/D converter 61 outputs input information with every sampling time interval Ts, and respective adders 62 add its outputs in time of 4×Ts.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a temperature sensor, and more particularly to an optical fiber type distributed temperature sensor.

[0002]

2. Description of the Related Art An optical fiber type distributed temperature sensor utilizes the fact that the intensity of scattered light such as Raman scattered light and Rayleigh scattered light in an optical fiber changes depending on the temperature, and this change is known as OTDR (Optical time Domain). The temperature distribution along the longitudinal direction of the optical fiber is measured by detecting with a method of Reflectometry).

An optical fiber type distributed temperature sensor utilizing Raman scattered light (hereinafter, simply referred to as Raman type temperature sensor)
The measurement concept of will be described below with reference to FIG.

When the pulsed light (pulse width Tw, pulse period Tp) is guided from the light source to the sensor optical fiber, backscattered light (reflected light) such as anti-Stokes light and Stokes light is excited in the optical fiber, Part returns to the measuring device. When this reflected light is measured at the sampling time interval Ts with the pulsed light incident time t = 0, the time functions Ia (t), Is (t) of the intensities of the anti-Stokes light and the Stokes light are the functions of the sampling time interval Ts. I want it. At this time, these ratios Ia (t) / Is (t) are purely a function of temperature, and after the light pulse is incident, the reflected light generated at the position of the distance X in the optical fiber is at the light pulse incident end ( Reflected light measurement unit)
2 × X / Co before returning to (C
o; the speed of light in the optical fiber), the linear temperature distribution along the optical fiber can be measured.

The time width Tr for measuring the reflected light is 2
× L / Co (L; optical fiber length), and the measured value in Tr gives effective temperature distribution information at this time.

Next, an outline of the Raman temperature sensor will be described with reference to FIG.

This Raman type temperature sensor has a measuring device 10
And the optical fiber 20 for sensor. When the pulsed light is guided from the light source 2 to the sensor optical fiber 20, the backscattered light (reflected light) is excited in the optical fiber, and a part of the excited reflected light returns to the measuring device 10 side, and the optical branching device is provided. 31,
It is guided to the optical branching device 32 via the optical fiber 22.

Of the reflected light divided by the optical splitter 32,
The light guided to the optical fiber 23a enters the anti-Stokes light OTDR measurement circuit 30a including the anti-Stokes light optical filter 4a, the light receiver 5a, and the averaging processing circuit 6a. The intensity time function Ia (t) is determined. On the other hand, the optical splitter 3
Of the backscattered light bisected by 2, the one that is guided to the optical fiber 23s is the Stokes light optical filter 4s,
The OTDR measuring circuit 30s for Stokes light composed of the light receiver 5s and the averaging processing circuit 6s is entered, and the time function Is (t) of the Stokes light intensity is obtained from this light intensity. The synchronization between the pulse light source 2 and the averaging processing circuits 6a and 6s is
The synchronization signal of the trigger circuit 1 is used, and the reflected light is sampled in the averaging processing circuits 6a and 6s at a constant time interval Ts shown in FIG.

The obtained time functions Ia (t) and Is (t) are input to the temperature distribution calculation circuit 7 and the calculation of Ia (t) / Is (t) is carried out, so that the optical fiber for the sensor is guided. The linear temperature distribution is measured.

As shown in FIG. 9, the averaging processing circuit 6 is composed of an A / D conversion circuit 61, an adder 62, a memory circuit 63 and a synchronizing circuit 64. The averaging process is performed as follows.

The analog quantity input from the light receiver 5 is A /
The digital amount is converted by the D conversion circuit 61, the sum of the digital amount and the digital amount stored in the memory circuit 63 is performed by the adder 62, and the result is again stored in the memory circuit 63.
To memorize. This operation is repeated for each pulse period TP, and the value finally stored in the memory circuit 63 is divided by the number of repetitions to obtain the average value of the input information. When this averaging process is performed, noise included in the input information is removed, so that the temperature measurement accuracy is improved.

Further, an A / D conversion circuit 61, an adder 62,
The synchronization of the memory circuit 63 is performed by the synchronization circuit 64.

In this Raman temperature sensor, for example, by laying an optical fiber for the sensor along the power cable, the temperature distribution in the longitudinal direction of the power cable can be known, and it can be used for controlling the power transmission capacity or the like. It is possible to detect a part where the temperature is high due to deterioration of the cable. Also, if it is used for fire detection in buildings, tunnels, etc., it is possible to locate the fire occurrence location.

[0014]

The Raman type temperature sensor or Rayleigh type temperature sensor is a promising method capable of measuring a linear temperature distribution by the above-mentioned method. Consideration is being made to improve it.

In order to improve the temperature accuracy, it is necessary to greatly increase the number of processing times of the averaging processing circuit in order to remove the influence of noise from the weak signal, and correspondingly, the averaging processing is performed. It is necessary to increase the number of processing bits of the circuit.

Further, in order to improve the distance resolution, it is necessary to shorten the sampling time.

However, as shown in FIG. 6, the processing time t becomes longer as the number of processing bits Nb is increased.
Even if a high-speed type circuit element is used, the averaging process may not be performed within the required sampling time interval Ts. In particular, this tendency becomes more remarkable as the sampling time interval Ts becomes shorter.

From this point of view, it has been difficult to improve the temperature accuracy and the distance resolution of the distributed temperature sensor.

SUMMARY OF THE INVENTION An object of the present invention is to solve the above problems and to provide an inexpensive optical fiber type distributed temperature sensor having high temperature accuracy and distance resolution.

[0020]

In order to achieve the above-mentioned object, the present invention provides a reflection pulse formed by backscattered light generated by causing an optical pulse to enter an optical fiber for sensor from a light source in a measurement system. The light is guided to the measurement system, the light intensities of these reflected lights are sampled and averaged by the averaging device, the temperature of the optical fiber is determined from the data, and the incident time of the optical pulse and the reflected light reach the measurement system. In the optical fiber type distributed temperature sensor for simultaneously measuring the temperature and the position by obtaining the generation position of the backscattered light from the time difference and measuring the temperature distribution of the optical fiber, the averaging processing device is A / D. It is configured by a converter and a plurality of sets of adder circuits that perform parallel processing of the output thereof, and the processing time of each set of adder circuits is within the number of sets times the sampling time interval.

With the above configuration, by processing the entire adder circuit in parallel by a plurality of circuits, even if the number of processing bits of the averaging processing circuit is increased, processing can be performed at short sampling time intervals, and temperature accuracy and distance resolution can be improved. It can be significantly improved.

[0022]

BEST MODE FOR CARRYING OUT THE INVENTION An embodiment of the present invention will be described in detail below with reference to the accompanying drawings.

First, for reference of the present invention, a reference example of an optical fiber type distributed temperature sensor utilizing Raman scattered light will be described with reference to FIG.

The basic concept and configuration of the optical fiber type distributed temperature sensor according to this reference example are almost the same as those of the conventional example shown in FIGS. 7 to 9, except that the addition circuit 600 of the averaging processing circuit 6 is different. Is divided into a pre-stage addition circuit 601 and a post-stage addition circuit 602.

At this time, the adder circuit 601 of the preceding stage adds the adder 62.
a and a memory 63a, and the post-stage addition circuit 602 is composed of an adder 62b and a memory 63b.

Next, the operation of the adder circuit 600 will be described.

The function of the adder circuit 600 is the same as that of the conventional one described with reference to FIG. 9, but the different points are as follows. That is, the value converted into a digital value by the A / D converter 61 is input to the pre-stage addition circuit 601, and each measurement input information is added here. When the number of additions reaches a certain number N0, the addition result is output to the subsequent stage. It is input to the adder circuit 602 and the memory in the memory 63a is cleared. The post-stage addition circuit 602 determines the number of additions N
For each 0, the addition result of the previous stage adder circuit 601 is input, added with the previously stored value, and input to the memory 63b. When this operation is repeated, the final result becomes the entire output of the adder circuit 600.

The number of times of addition N0 may be set to a constant value, or may be a stage at which the addition result of the preceding stage addition circuit 601 has reached the processing bit or more.

Next, a method of selecting the processing bit number Nb of the pre-stage addition circuit 601 will be described. Here, the sampling time interval Ts is Ts = 20 nS, the A / D converter 6
The processing bit number of 1 is 8 bits, and the post-stage addition circuit 602
The case where the number of processing bits of is set to 32 will be described.

The processing time of the pre-stage addition circuit 601 and the post-stage addition circuit 602 (hereinafter referred to as "pre-stage" and "post-stage" as necessary) is t.
Let a and tb, and let the respective processing allowable times be Ta and Tb.

Since the processing time ta of the preceding stage increases linearly according to the number of processing bits Nb, if the proportional constant (slope) is set to b and the delay when Nb = 0 is set to a constant a. , The processing time ta of the preceding stage is expressed by the following equation.

Ta = a + bNb Further, since the processing time tb in the subsequent stage has nothing to do with the Nb to be obtained, this is set as a constant tb0.

Tb = tb0 Next, the processing of the previous stage is performed by sampling time interval Ts (20n
Since it is possible only within S), the processing allowable time Ta of the preceding stage is determined by the sampling time interval Ts.

Ta = Ts Further, the processing allowable time Tb in the subsequent stage is equal to the memory 63a in the previous stage.
Considering that it is not necessary to operate the subsequent stage until the maximum value becomes, the following relationship is established with respect to the sampling time interval Ts.

Tb = Ts.multidot.2 ^{Nb-8 In} this way, the processing allowable time Tb of the subsequent stage becomes longer than the processing allowable time Ta of the preceding stage because it is not necessary to operate the subsequent stage until the memory 63b of the preceding stage becomes maximum. The minimum time is determined when the input to the A / D converter 61 reaches the maximum value (8 bits) each time, and the ratio is 2 ^Nb / 2 ⁸ = 2.
^Because it will be ^Nb-8 .

A point to be taken into consideration is that since the processing allowable time Tb of the post-stage addition circuit 602 is relatively long, it is advantageous to reduce the number of processing bits Nb of the pre-stage addition circuit 601.
Furthermore, it is a device that does not require the use of high-speed elements for the elements used.

Here, the ratios of the processing times ta and tb of the front and rear stages and the processing permissible times Ta and Tb are calculated, and the ratios of the pre-processing aptitude ka and the post-processing aptitude index kb are respectively calculated.
When you place a, a ka = ta / Ta = (a + bNb) / Ts kb = tb / Tb = (tb0 / Ts) · 2 8/2 Nb ... (1).

FIG. 2 shows the processing bit number Nb in the preceding stage and the above (1) when the processing time tb in the latter stage is used as a parameter.
An example of the relation of expressions is shown. When the number of processing bits Nb in the preceding stage is increased, the preprocessing aptitude index ka rises linearly and the postprocessing aptitude index kb conversely decreases exponentially, as can be estimated from the equation (1). In this FIG.
It is preferable that both the pre-stage process suitability indexes ka and kb are 1 or less and close to 1.

Regarding the post-stage processing suitability index kb, when the number of processing bits Nb in the preceding stage is small, for example, 8 bits which is equal to the number of processing bits of the A / D converter 61, a high-speed circuit with a processing time of about tb = 30 ns. Even when using the element,
Since the processing suitability index kb of the latter stage is less than 1 and there is no room for the latter stage processing, it becomes necessary to use higher speed circuit elements. On the other hand, if the number of processing bits Nb in the preceding stage is increased to 32 bits, the demand for the speed of the element to be used becomes loose, but the processing suitability index ka in the preceding stage exceeds 1, and there is no room for the processing in the preceding stage. In summary, the following can be concluded from this FIG.

(1) Number of processing bits Nb of the pre-stage addition circuit 601
Between the number of processing bits of the A / D converter 61 (8 bits) and the number of processing bits of the subsequent stage (32 bits), both process suitability indexes ka and kb are obtained when 32 bits of 30 ns processing elements are used. It becomes smaller than the value (* mark in the figure) corresponding to the processing time of.

(2) Processing suitability index ka of the pre-stage addition circuit 601
Becomes smaller as the number of processing bits Nb becomes smaller,
On the contrary, the processing suitability index kb of the post-stage addition circuit 602 becomes smaller as the pre-stage processing bit number Nb becomes larger. However, since the post-stage process suitability index kb exponentially decreases with respect to the pre-stage process bit number Nb, the effect is large only by making Nb slightly larger than the process bit number of the A / D converter 61.

For example, if the number of processing bits Nb of the pre-stage addition circuit 601 is 16 bits, the processing suitability index ka = 0.
It becomes 8 and is within 1 and the processing time ta is 16n.
s and the target sampling time interval Ts = 20 ns.

When the post-processing suitability index kb is set to the same value as the pre-processing suitability index ka, the post-processing processing time is tb.
= 4 (16 ns × 2 ⁸ ), which is sufficiently slow as a 32-bit processing element of the post-stage adder circuit.

FIG. 1 shows an embodiment of the present invention in which the above-mentioned adder circuit 600 includes a latch circuit 65 and four adders 62.
And a memory 63 so that the processing time of each set can be dealt with by four times the sampling time Ts.

The A / D converter 61 outputs the input information at every sampling time interval Ts, and the result thereof is latched by the latch circuit 6.
5 and inputs the output of the latch circuit 65 to each adder 62.
Is to be added within the time of 4 × Ts. Conversely, if circuit elements having the same function are used, the sampling time interval Ts can be shortened to 1/4. In this embodiment, the adder 62
Although four sets are used, the number of sets can be arbitrarily selected.

FIG. 4 is a combination of the technique shown in FIG. 1 and the technique shown in FIG. 3. The processing time of the adder circuit 600 is significantly shortened by using the parallel adder type as the pre-stage adder circuit 601. It is possible.

FIG. 5 shows the minimum practicable sampling time intervals actually measured by the circuit configuration of the present invention or the reference example using the circuit elements having the same function. I know I can handle the time.

In the above embodiment, the addition circuit was taken into consideration, but it goes without saying that the same effect can be obtained also by configuring the A / D converter as a parallel adder type.

[0049]

The present invention exhibits the following excellent effects.

(1) Even if a circuit component having the same function as the conventional one is used, it is possible to realize an averaging processor having a high processing bit number and a short sampling time.

(2) Since the number of processing bits can be increased, the number of times of averaging processing can be increased and the influence of noise can be eliminated.
As a result, an optical fiber type distributed temperature sensor with high temperature accuracy can be realized.

(3) Since the sampling time can be shortened, an optical fiber type distributed temperature sensor having a high distance resolution can be realized.

(4) Since it is not necessary to develop a new circuit component, a high-performance device can be realized at low cost.

[Brief description of drawings]

FIG. 1 is a configuration diagram of an averaging processing circuit of an optical fiber type distributed temperature sensor showing an embodiment of the present invention.

FIG. 2 is a diagram showing a relationship between the number of processing bits of a pre-stage addition circuit and a processing suitability index.

FIG. 3 is a configuration diagram of an averaging processing circuit showing a reference example of the averaging processing circuit.

FIG. 4 is a configuration diagram of an averaging processing circuit showing an application example of the present invention.

FIG. 5 is an explanatory diagram comparing the performance of the present invention with a conventional type and a reference example.

FIG. 6 is a diagram showing the relationship between the number of processing bits and processing time.

FIG. 7 is a diagram showing a measurement concept of a conventional optical fiber type distributed temperature sensor.

FIG. 8 is a block diagram of an optical fiber type distributed temperature sensor that has been conventionally considered.

FIG. 9 is a configuration diagram of a conventional averaging processing circuit.

[Explanation of symbols]

 61 A / D conversion circuit 62, 62a, 62b adder 63, 63a, 63b memory 64 synchronization circuit 65 latch circuit 600 addition circuit 601 pre-stage addition circuit 602 post-stage addition circuit

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Hisaichi Sasahara 5-1-1 Hidaka-cho, Hitachi-shi, Ibaraki Hitachi Cable, Ltd.

Claims (1)

[Claims] 1. An optical pulse is made incident on a sensor optical fiber from a light source in the measurement system, reflected light formed by backscattered light generated in the fiber is guided to the measurement system, and the light intensities of these reflected lights are averaged. By averaging and sampling by the data processing device, by determining the temperature of the optical fiber from the data, by determining the generation position of the backscattered light from the difference between the incident light time of the optical pulse and the time when the reflected light reaches the measurement system, In an optical fiber type distributed temperature sensor for simultaneously measuring temperature and position and measuring the temperature distribution of the optical fiber, the averaging processor is an A / D converter and a plurality of sets of adder circuits for parallel processing the outputs thereof. And an optical fiber type distributed temperature sensor, wherein the processing time of the adder circuit of each set is within the number of times of the sampling time interval.